Zung WWK. A Rating Instrument For Anxiety Disorders. Psychosomatics.
Zung WWK. A Rating Instrument For Anxiety Disorders. Psychosomatics.
PSYCHOSOMATICS
OFFICIAL JOURNAL OF THE ACADEMY OF PSYCHOSOMATIC MEDICINE
• Diagnosis and diagnostic criteria enable Diagnostic and Statistical Manual, Second
clinicians and researchers alike to com- Edition or DSM-II as a neurosis characterized
municate with one another about their pa- by anxious overconcern extending to panic
tients and make comparisons of cases and and frequently associated with somatic symp-
data meaningful. The fact there is a need for toms 2 • More detailed definitions and descrip-
assessing anxiety, whether as an affect, a tions of anxiety and its characteristic symp-
symptom, or a disorder is obvious by the toms are described by the following authors
numerous rating scales available and in use. and summarized in Table I: Kolb in Noyes'
However, the need for a standardized method Modern Clinical Psychiatry3, Lief in Com-
of evaluating and recording the presence of prehensive Textbook of Psychiatry" Portnoy
anxiety as a clinical disorder has not been in the American Handbook of Psychiatry l, and
met by most scales today. We were interested a report on anxiety neurosis by Wheeler,
in having a rating instrument which would White, Reed and Cohen5 •
fulfill the following: it should be inclusive with In devising our rating instrument, diag-
respect to symptoms of anxiety as a psychiat- nostic criteria used were comprised of the most
ric disorder, it should quantitate the symp- commonly found characteristics of an anxiety
toms, it should be short and simple, and it disorder, such as those listed in Table I. From
should be available in two formats so that, 1) this list of criteria, an observer rated Anxiety
the patient can indicate his own responses on Status Inventory (ASI) and a patient Self-
a self-administered scale, and 2) the observer rating Anxiety Scale (SAS) were constructed.
can indicate his clinical evaluation of the pa-
Anxiety Status Inventory (ASI)
tient's status on the same set of criteria. This
report summarizes our efforts at devising a Table II is the form of the ASI which is
rating instrument for anxiety disorders which the clinician rated instrument. It contains the
we felt fulfilled our needs and purposes. diagnostic criteria for anxiety as a psychiat-
ric disorder (5 affective and 15 somatic symp-
METHOD
toms) and the interview guide for eliciting
In the construction of the present rating each of the symptoms. The data upon which
instrument the symptoms of the illness were the judgments are based come from the inter-
delineated by using the descriptive approach, view with the patient. The items in the scale
since the basis of definition and classification are to be quantified by using all of the infor-
in psychiatric nosology continues to be based mation available to the rater. This includes
upon presenting symptomatologyl. both clinical observations and the material
Anxiety as a disorder is defined in the reported by the patient.
Dr. Zung is Associate Professor of Psychiatry, Use of the Interview Guide assures cover-
Duke University Medical Center, and Veterans age of all of the areas in which judgments are
Administration Hospital, Durham, N.C. required. However, the rater has the flexibility
SO~IATIC SY:IPTOMS
Musculoskeletal $vstem
Tremor, ~uscle tightness, Increased tension, Trel:lbling 53.5
tremors, spasms, tremors, stiffn"ss Shakiness 46.5
painful movements,
Tension headache. ileadaches, neck and Headache 58.3
back pains Tires easil~.. 95.0
h'~akness, h'eakness, Weakness 56.0
k",stlessgess Restlessness Fatigued all
the time 45.1
I'!!.d.!o~'!.s cUl-,!,--S ys t em
l'alpitation, Palpitation, throbbing Palpitation, rapid pulse, Palpitation 96.7
rapid heartbeat pain in chest increased Bp
B-e_,~p.iJ-,,-ton'
Sys tem
Hyperv",ntilation: Rapid or irr~gu13r
dizziness, Dizziness, breathing Dizziness 78.3
fainting, Faintness 70.0
shortne::;s of breatil, :::;lll.)rtness of ureath) llreathlessness 90.0
f",eling of choking, Lvnstriction in chest, Breath
pressure on tilorax, unsatisfactpry 52.7
parestllesias Paresthesias 58.2
~L:n_i_t.9-u_~~.!!...a_ ~~ t ~~~.
Incre3sed desire to rrillary frequellcy, Urinary frequenc~' Frequency 18.6
urinate urgcIH':y
Skin
rac,", f 1uSiled, flushing of f:lce, Flushing or ~3llor, Flushing 36.2
perspiration Sensation of heat cold, wet extremities Sweating 44.9
system, taking into account Severity in terms part of the time in frequency
of: intensity, duration, and frequency. These 4 severe in intensity or duration, pre-
are defined as follows: sent most or all of the time in fre-
1 = none or insignificant in intensity or quency.
duration, present none or a little of To help establish severity, the following
the time in frequency questions may be necessary: Intensity -
2 - mild in intensity or duration, present ''How bad was it?" Duration - "How long
some of the time in frequency did it last?" Frequency - ''How much of the
3 - of moderate severity, present a good time did you feel that way?"
.', . Mental disintegratior 00 you ever fee I I ike you1re falling apart? I 2 J .',
Going to pieces:
H~lvcyou ever felt uneasy? or that something
5. Apprehension terrible was going to happen? 1 2 J .',
7. Body aches '-. pains Do you have headaches? neck or back pains? I 2 3 .',
I.',. Paresthesias
Ever have feelings of numbness and t ing ling
in your fingertips? or around your mouth? I 2 J .',
IS. Nausea & vomiting Do you ever feel sick to your stomach or I 2 3 .',
feel like vomiting?
16. Urinary frequency How often do you need to empty your bladder? I 2 3 4
18, Face flushing Do you ever feel your face getting hot and 1 2 3 4
blushing?
19. Insomnia How have you been sleeping? 1 2 3 4
4. An items is scored positive and present c. Patient admits that symptom is still a
when: problem
A. Behavior is observed 5. An item is scored negative and not pre-
B. Behavior was described by the patient sent when:
as having occurred A. Symptom has not occurred and not a
None OR :
A little Some Good part Most OR AlII
of the time of the time of the time of the time
1- I feel more nervous and
anxious than usual
7. I am bothered by he~daches,
neck and ba~k pains
None OR
A little Some Good part Most OR All
of the time of the time of the time of the time
1- I feel more nervous and
anxious than usual I 2 3 :.
2. I feel afr,1id for no reason at all 1 2 J :.
J. I i!et upset e3si Iy or fee 1 pnnicky 1 2 J .:.
4. I fee 1 like I'm fctllin~ apart and 1 :2 J -,
going to pieces
5. I fee 1 that everythi ng is all right ~ 3 :2 1
and nothing bad wi 11 happen
6. My arms and legs shake and tremble I :2 J .:.
7. I am bothered by headaches, 1 2 3 .:.
neck 3nd bilck pilins
so that of the 20 items used, some of the items a higher score. In scoring the SAS, a value of
were worded symptomatically positive, and 1, 2, 3 and 4 is assigned to a response depend-
others symptomatically negative, depending ing upon whether the item was worded posi-
upon their suitability and usage. In addition, tively or negatively. A key for scoring this
an even-number of columns were used to scale can be simply made up, as shown in
eliminate the possibility of a patient checking Table IV.
middle and extreme columns. An index for the SAS (and for the ob-
In using the scale, the patient was asked server rated Anxiety Status Inventory or
to rate each of the 20 items as to how it ap- ASI) was derived by dividing the sum of the
plied to him within the past week, in the values (raw scores) obtained on the 20 items
following four quantitative terms: None by the maximum possible score of 80, con-
OR A little of the time, Some of the time, verted to a decimal and multiplied by 100
Good part of the time, Most OR All of the (see Table V).
time. The SAS is constructed so that the less In order to prevent confusion between the
anxious patient will have a low score on the ASI and SAS results, the converted ASI score
scale, and the more anxious patient will have is called a Z score and the converted SAS
TABLE V TABLE VI
A table for the conversion of raw scores to Distribution of patients tested by their ase,
ASI and SAS indices sex, and psychiatric diallnoses at time of
discharse
ASI ASI ASI
& & & AGE N %
Raw SAS Raw SAS Raw SAS
19 & under 8 3.6
Score Index Score Index Score llIdex
20 - 39 97 43.1
20 25 40 50 60 75 40 - 64 112 49.7
21 26 41 51 61 76 65 & over 8 3.6
22 28 42 53 62 78
225 100.0
23 29 43 54 63 79 SEX
24 30 44 55 64 80 Males 175
25 31 45 56 65 81 Females 50
26 33 46 58 66 83
225
27 34 47 59 67 84
DIAGNOSES
28 35 48 60 68 85
Mental Retardation 1 0.4
. 29
30
36
38
49 61 69 86
Organic Brain Syndrome 5 2.2
50 63 70 88 Schizophrenia 25 11.1
31 39 51 64 71 89 Anxiety Disorder 22 9.8
32 40 52 65 72 90 Depressive Disorder 96 42.9
33 41 53 66 73 91 Obsessive Compulsive Disordel' 7 3.2
34 43 54 68 74 92 Personality Disorders 54 24.2
35 44 55 69 75 94 Psychophysiologic Disorders 2 0.9
36 Transient Situational
45 56 70 76 95
Disturbances 12 5.3
37 46 57 71 77 96
38 48 58 73 78 98 100.0
39 49 No Psychiatric Diagnosis 1
59 74 79 99
80 100 225
score is called an index. The self-rated forms were not looked at nor
scored until the completion of the study.
Taylor Manifest Anxiety Scale (TMAS)
The SAS was given on randomly selected
The TMAS consists of 50 items drawn
days to a normal control group of 100 in-
from the Minnesota Multiphasic Personality
dividuals, who were at work at the hospital.
Inventory (MMPI) judged to be indicative
They consisted of approximately an equal
of manifest anxiety.6 Since this is a commonly
number of professional and non-professional
used scale in anxiety studies, we included it
staff members.
in our investigation.
RESULTS
Data CoIIection
Subjects Tested
The data for this study were obtained as
follows: all new patients admitted to the psy- A total of 225 patients were tested during
chiatric in-patient service of the hospital for the period of study. The inpatient population
15 consecutive months, and all new patients (N=152) were all men, whose ages ranged
seen at the out-patient clinic for 4 consecutive from 22 to 75 years (m=45). The out-pa-
months were seen and tested. Patients were tient population (N=73) had 23 men and
first given the self-rating form of the anxiety 50 women, whose ages ranged from 14 to 72
scales (SAS and TMAS), after which an in- years old (m=32). The total mean age for
terview was conducted to complete the inter- all 225 subjects was 41 years old. Each pa-
view-rating form of the anxiety scale (ASI). tient was given a diagnosis at the time of dis-
charge by his psychiatrist, and was done
TABLE VII without any knowledge of the scale results.
The distribution of all patients tested by their
Anxiety Status Inventory (ASI) Z Scores for age, sex and diagnoses are presented in
the various diagnostic groups tested Table VI.
ASI Z Score A total of 100 normal adult subjects were
Group Diagnosis N Mean & S.D. tested. There were 57 men and 43 women.
1 Anxiety Disorder 22 62.0
Their ages ranged from 18 to 62 years old
± 13.8
2 Schizophrenia 25 49.4 ± 15.9
(m=34).
3 Depressive Disorder 96 49.9 ± 12.5 Anxiety Status Inventory (ASI)
4 Personality Disorder 54 52.6 ± 13.6
Results of the ASI for the five largest diag-
5 Transient Situational
Disturbances 12 42.0 + 8.1
nostic groups which comprised 93% of the
total patient population tested are presented
TABLE VIII in Table VII. Statistical tests of significance
using analysis of variance indicated that the
Self-rating Anxiety Scale (SAS) indices for mean ASI Z score obtained by patients with
the various diagnostic groups tested diagnoses of anxiety disorders was significant-
ly higher than those of the other four diag-
SAS Index
nostic groups (P = < 0.05).
Grou,p Diagnosis N Mean & S.D.
Self-rating Anxiety Scale (SAS)
1 Anxiety Disorder 22 08.7 ± 13.5
2 Schizophrenia 25 46.4 ± 12.9 Results of the SAS for the five largest diag-
3 Depressive Disorder 96 50.7 ± 13.4 nostic groups are presented in Table VIII.
4 Personality Disorder 54 51.2 ± 13.2 Analysis of variance indicated that the mean
5 Transient Situational SAS index obtained by patients with diag-
Disturbances 12 45.8 ± 11.9 nosis of anxiety disorders was significantly
higher than those of the other four diagnostic
209 groups (P = < 0.05). In addition, the mean
6 Controls 100 33.8 ± 5.9 SAS index obtained from normal control sub-
jects was significantly lower than all five of as being the worst (upper third) the follow-
the patient diagnostic groups (P = < 0.01). ing affective symptoms: feelings of mental
Table IX rank orders the 20 items of the disintegration, anxiousness and apprehension,
SAS as quantitated by the patients with anxi- and somatically: symptoms referable to the
ety disorders, listed in decreasing order of musculoskeletal and gastrointestinal systems.
severity, and arbitrarily divided into thirds. It By comparison, Wheeler et al.5 noted in their
can be seen that these patients complained study that the most frequent complaints by
patients with anxiety neurosis as involving
TABLE IX the cardiovascular and respiratory systems
(see Table I).
Severity of Symptoms of Patients with
Anxiety Disorders Taylor Manifest Anxiety Scale (TMAS)
Results of the TMAS for the five largest
Mean
Rank SAS Item Items: in Decreasing diagnostic groups are presented in Table X.
Order Score No. Order of Severity Analysis of variance indicated the mean
TMAS scores obtained for these diagnostic
1 2.8 4 Mental disintegration
groups were not significantly different from
2.8 6 Tremors
each other (P = > 0.05).
2 2.6 7 Body aches and pains
3 2.5 1 Anxiousness Correlation between ASI, SAS and T M AS
2.5 5 Apprehension Pearson product-moment correlation for
2.5 15 Nausea and vomiting calculation of the coefficient r on data ob-
4 2.4 2 Fear tained from all patients was performed. All
2.4 3 Panic of the coefficients r calculated were statisti-
2.4 10 Palpitation cally significant, with P = < 0.01 in all in-
2.4 14 Paresthesias stances.
2.4 16 Urinary frequency The correlation between the ASI and SAS
5 2.2 17 Sweating was 0.66. Correlation between the ASI and
6 2.1 8 Fatigue TMAS, and SAS and TMAS were 0.33 and
2.1 13 Dyspnea
0.30, respectively. Correlation between the
2.1 20
ASI and SAS scores for patients with a diag-
Nightmares
7
nosis of anxiety disorder was 0.74.
2.0 9 Restlessness
Split half correlations for the ten even-
2.0 11 Dizziness
numbered and the ten odd-numbered ASI
2.0 12 Faintness
items, and similarly, for the even-odd SAS
2.0 18 Face Blushing
items were 0.83 and 0.71, respectively.
2.0 19 Insomnia
Correlations between items 1 through 20
of the ASI with the ASI Z score were as fol-
TABLE X lows: 0.50, 0.56, 0.65, 0.65, 0.64, 0.50, 0.52,
0.34, 0.51, 0.65, 0.61, 0.65, 0.42, 0.63, 0.60,
Taylor Manifest Anxiety Scale (TMAS) 0.39, 0.54, 0.49, 0.58, and 0.47, respectively.
Scores for the various diagnostic groups tested Correlations between items 1 through 20
TMASScore
of the SAS with the SAS index were as fol-
Group Diagnosis N Mean & S.D. lows: 0.39, 0.53, 0.57, 0.69, 0.50, 0.56, 0.58,
1
0.40, 0.50, 0.64, 0.47, 0.51, 0.27, 0.55, 0.61,
Anxiety Disorder 22 31.0 ± 15.3
0.62,0.42, 0.47, and 0.49, respectively.
2 Schizophrenia 25 23.0 ± 13.5
3 Depressive Disorder 96 25.1 ± 13.5 DISCUSSION
4 Personality Disorder 54 31.7 ± 10.8 Anxiety scales available and in use today
5 Transient Situational can be divided into those which are general
Disturbances 12 29.3 ± 7.0 and measure anxiety as a personality trait or
feeling state, and those which are specific and or as a self-rated scale (Self-rating Anxiety
measure anxiety as a clinical entity. They may Scale or SAS). Data was collected from a
also be grouped as those which are self-ad- population of psychiatric patients using the
ministered and those which are given by a new instrument and the Taylor Manifest
trained interviewer. General instruments in- Anxiety Scale (TMAS). Statistical analyses
clude those by Costello and Comrey;, whose of the results indicated that the new instru-
anxiety scale was designed to measure a pre- ment was able to differentiate significantly
disposition to develop anxious states; Cattell's anxiety patients from patients with other diag-
anxiety scale" which is used to measure fluc- noses, whereas the TMAS did not. Correla-
tuations in level of anxiety over short periods tion between the ASI and SAS, and between
of time; and Taylor's manifest anxiety scaler,. the individual items of the two-part instru-
Use of a general scale for the measurement of ment with their respective total scores were
anxiety as a clinical entity may tend to obscure all significant.
results more than it would reveal. Thus, in a REFERENCES
study of the TMAS, O'Connor et aP did a
1. Portnoy, I.: The Anxiety States, in Arieti, S.
factor analysis of their results and showed (Editor): American Handbook of Psychiatry,
that as many as five different dimensions were Vol. I: Basic Books, New York, 1959, p. 308.
involved in the items comprising the scale. 2. Diagnostic and Statistical Manual of Mental
Aggregate scores of such a scale may confound Disorders, Second Edition, American Psy-
by blurring, and analysis by sub-groups of chiatric Association, Washington, D.C., 1968,
p.39.
items may fail to indicate differences which 3. Kolb, L.: Noyes' Modern Clinical Psychiatry,
may be present. Examples of specific anxiety Seventh Edition, Saunders, Philadelphia, 1968,
scales include those by Hamilton10 and Kell- pp. 464-466.
ner and Sheffield l l • These are both interview- 4. Lief,. H.: Anxiety React;on, in Freedman, A.
er-rated scales and require a trained clinician and Kaplan, H. (Editors): CcmLIJrehensive
Textbook of Psychiatry, Williams and Wil-
to complete them.
kins, Baltimore, 1967, pp. 865-866.
In developing the presently reported rat- 5. Wheeler, E., White, P., Reed, E., and Cohen,
ing instrument for measuring anxiety as a M.: Neurocirculatory Asthenia (Anxiety Neu-
clinical entity, we felt that it would fullfill the rosis, effort syndrome, neurasthenia), JAMA,
previously mentioned criteria and that it 142: 878-888, 1950.
6. Taylor, J.: A personality scale of manifest
would be a useful tool for our purpose. From
anxiety, J. Abn. & Soc. Psychol. 48:285-290,
the results obtained, it did appear to us that 1953.
the two-part instrument, consisting of the self- 7. Costello, C. and Comrey, A.: Scales for mea-
rated SAS, and the clinician-rated ASI was suring depression and anxiety, J. Psychol.
useful in quantitating anxiety as an opera- 66 :303-313, 1967.
8. Cattell, R.: Handbook for the IPAT Anxiety
tionally defined disorder.
Scale. Institute for Personality and Ability
SYNOPSIS Testing, Champaign, Illinois, 1957.
9. O'Connor, J., Lorr, M. and Stafford, J.: Some
A rating instrument for the measurement patterns of manifest anxiety, J. Clin. Psychol.
of anxiety as a clinical entity was devised as 12: 160-163, 1956.
an attempt to quantitate the symptoms of this 10. Hamilton, M.: The assessment of anxiety
disorder, using the most commonly agreed states by rating, Brit. J. M ed. Psychol. 32:
50-55, 1959.
upon diagnostic criteria. As a two-part instru- 11. Kellner, R. and Sheffield, B.F.: Symptom Rat-
ment, it can be used as an interviewer-rated ing Test scores in neurotics and normals,
inventory (Anxiety Status Inventory or ASI), Brit. J. Psychiat. 113 :525-526, ]967.